Cloud attenuation stochastic characterization from ground-based microwave radiometric data at Ka-band

This work aims at characterizing the behavior of suspended clouds in terms of atmospheric path attenuation. Well-known radiative models are adopted to provide an estimate of the equivalent clear-air path attenuation contribution, exploiting surface weather measurements and making several assumptions on their vertical stratification over the troposphere. However, the attenuation contribution due to non-precipitating clouds cannot be easily modelled by only using in-situ measurements, i.e., surface boundaries are not able to provide enough information about the whole atmospheric status for a given instant. A stochastic approach is used to model the time evolution of the cloud contribution. Both the probability density function and the power spectral density are retrieved by exploiting measurements from the RPG-HATPRO radiometer installed in Cebreros, Spain at the European Space Agency's Deep Space Antenna site. Physically-based prediction models for all-weather path attenuation estimation at 32 GHz are applied to the measured radiometric brightness temperatures. The cloud contribution is then extrapolated and modelled as a log-normal stochastic process as a result of a detailed analysis in both amplitude and time domains. Observation continuity is the key to improve the long-term statistical characterization of the atmospheric behavior and such modelling has proved to be crucial in assessing and completing attenuation estimate datasets, whenever solid path attenuation estimates were not available for relatively long periods.